Bug 1881245

+++ This bug was initially created as a clone of Bug #1881244 +++

Description of problem:

When discovering many target portals in a single SendTargets discovery command, iscsiadm will crash due to a strcat buffer overflow.

Version-Release number of selected component (if applicable):

iscsi-initiator-utils > 6.2.0.878-0

How reproducible:

100%

Steps to Reproduce:

1. Configure an iSCSI target to report > 85 targets
   (you can just create a bunch of target names with targetcli, they don't need LUs and ACLs set)

2. iscsiadm -m discovery -t set -p <discovery portal>

Actual results:

iscsiadm will crash 

Expected results:

all targets discovered and node records created

Additional info:

This is a regression introduced with the new CHAP modes, a buffer in a struct that gets reused for each target discovered isn't reset and is appended to with the chap_algs config string with each pass until it overflows.

Reported upstream (https://github.com/open-iscsi/open-iscsi/issues/219) by a Fedora user.  Fix verified by that user, applied upstream, and to Fedora.

One line fix:
--- a/usr/idbm.c
+++ b/usr/idbm.c
@@ -169,6 +169,7 @@ static struct idbm *db;
 #define __recinfo_int_list(_key,_info,_rec,_name,_show,_tbl,_n,_mod) do { \
        _info[_n].type = TYPE_INT_LIST; \
        strlcpy(_info[_n].name, _key, NAME_MAXVAL); \
+       _info[_n].value[0] = '\0'; \
        for (unsigned long _i = 0; _i < ARRAY_LEN(_rec->_name); _i++) { \
                if (_rec->_name[_i] != (unsigned)~0) {                  \
                        for (unsigned long _j = 0; _j < ARRAY_LEN(_tbl); _j++) {        \